Research
- Effects of item and unit nonresponse in surveys (MCAR, MAR and NMAR), estimation methods for complex samples
- Congeniality
- Data fusion and identification
- Evaluation of statistical matching
- Questionnaire splitting to reduce respondent burden
- Data anonymization to generate scientific-use-files
- Econometric estimation of productivity using establishment survey data
- Evaluation research (especially of education measures) and the problem of self-selection
- Statistical Sampling
- Markov Chain Monte Carlo methods
- Bayesian statistics